Super-efficient Echocardiography Video Segmentation via Proxy- and Kernel-Based Semi-supervised Learning
Authors: Huisi Wu, Jingyin Lin, Wende Xie, Jing Qin
AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments have been conducted on two famous public echocardiography video datasets, Echo Net-Dynamic and CAMUS. Our model achieves the best performance-efficiency trade-off when compared with other state-of-the-art approaches, attaining comparative accuracy with a much faster speed. |
| Researcher Affiliation | Academia | Huisi Wu1*, Jingyin Lin1, Wende Xie1, Jing Qin2 1 College of Computer Science and Software Engineering, Shenzhen University 2 Centre for Smart Health, The Hong Kong Polytechnic University |
| Pseudocode | No | The paper does not contain structured pseudocode or algorithm blocks. |
| Open Source Code | Yes | The code is available at https://github.com/Jingyin Lin/PKEcho-Net. |
| Open Datasets | Yes | We evaluated our method on two public echocardiography video datasets: Echo Net-Dynamic (Ouyang et al. 2020) and CAMUS (Leclerc et al. 2019) datasets. |
| Dataset Splits | Yes | We split the training set, validation set, and test set with a ratio of 7:1:2, where four kinds of data augmentations are used to enrich the video data diversity for training |
| Hardware Specification | Yes | Efficiency comparison with the state-of-the-art methods on one RTX 3090 GPU at 320 x 320 resolution. |
| Software Dependencies | No | The paper states 'We implemented our method with the Py Torch framework' but does not specify version numbers for PyTorch or any other software dependencies. |
| Experiment Setup | Yes | We trained our model for 50 epochs with a poly strategy, where the learning rate is multiplied by (1 iter itermax )0.9 for each iteration with an initial learning rate of 1e-3 for all experiments. We set batchsize = 8 and an Adam optimizer (Kingma and Ba 2014) is also used to accelerate the convergence. |